Reconfigurable logic structures

Abstract

Reconfigurable electronic structures and circuits using programmable, non-volatile memory elements. The programmable, non-volatile memory elements may perform the functions of storage and / or a switch to produce components such as crossbars, multiplexers, look-up tables (LUTs) and other logic circuits used in programmable logic structures (e.g., (FPGAs)). The programmable, non-volatile memory elements comprise one or more structures based on Phase Change Memory, Programmable Metallization, Carbon Nano-Electromechanical (CNT-NEM), or Metal Nano-Electromechanical device technologies.

Description

CROSS-REFERENCE TO RELATED PATENT APPLICATIONS[0001]This application relates to and incorporates by reference co-pending and commonly assigned U.S. patent application Ser. No. 11 / 267,788 entitled “Multi-Terminal Phase Change Devices” and to co-pending and commonly assigned U.S. patent application Ser. No. 11 / 267,789 entitled “Methods for Fabricating Multi-Terminal Phase Change Devices”.FIELD OF THE INVENTION[0002]The present invention relates to programmable logic devices and other programmable structures. More specifically, the present invention relates to programmable logic structures using phase change material devices and other non-volatile, reconfigurable device technologies.BACKGROUND OF THE INVENTION[0003]Digital electronic systems are comprised of essentially three different types of components: memory, microprocessor, and logic components. Memory devices store information such as, for example, the contents of a spreadsheet or database. Microprocessors execute software instr...

AI Technical Summary

Benefits of technology

"The patent describes using programmable, non-volatile memory elements to create electronic structures and circuits. These memory elements can perform the functions of storage and a switch to produce components used in programmable logic structures. The memory elements can be based on technologies such as Phase Change Memory, Programmable Metallization Device, Carbon Nanotube Nano-Electromechanical, or Metal Nano-Electromechanical devices. The invention allows for the creation of highly programmable and flexible electronic structures and circuits."

Problems solved by technology

Once manufactured, these functions cannot be altered.

Early programmable logic device architectures, including the technologies used to implement their programmability, were often not up to par with their application specific integrated circuit (ASIC) counterparts.

Anti-fuse, EEPROM / FG and SONOS NVM technologies also have a limited lifespan, since they can be reprogrammed only a finite number of times; require special charge pump circuitry to generate the high voltages needed for programming; and are difficult to scale with CMOS scaling.

These drawbacks, particularly when weighed against the benefits gained by use of SRAM technology has led SRAM to become the leading programmable switch technology used in FPGAs.

Although SRAM-based programmable switch technology has become the preferred programmable switch technology, its use does present other drawbacks.

One major drawback of SRAMs is that they occupy a large percentage of the programmable fabric of an SRAM-based FPGA.

Hence, even a single 6T SRAM configuration bit occupies a significant amount of base silicon.

This problem is compounded by the fact that present day SRAM-based FPGAs can contain ten to fifty million SRAM cells.

Accordingly, SRAM-based FPGAs, although offering many benefits, have the serious drawback of the SRAM cells consuming a large portion of the FPGA chip area.

SRAMs are also susceptible to radiation-induced soft errors.

A radiation-induced soft error occurs when neutrons or alpha particles from the environment impinge on the SRAM and cause it to change state.

Finally, SRAM-based FPGAs are volatile (see FIG. 4), meaning that the SRAM-based configuration bits must be reprogrammed (i.e. reconfigured) every time the FPGA is powered down and then powered up again.

In addition to the drawback of having to wait for the FPGA to be configured to boot-up, large amounts of power are needed to complete the boot up process.

The need for special power supply and control circuitry adds further complexity and cost to systems using SRAM-based FPGAs.

Images